Introduction and objectives: Information comparing left atrial appendage closure (LAAC) to direct oral anticoagulation (DOAC) therapy is scarce. Our aim is to compare the clinical outcomes between LAAC and DOACs on an elderly population (> 80 years of age).

Methods: We retrospectively collected 1144 octogenarian patients with atrial fibrillation from 3 different tertiary hospitals. A total of 970 patients received DOACs and 174 patients were treated with LAAC. At baseline, both groups had similar cardiovascular risk factors. The LAAC group had more history of bleeding, anemia or previous cancer. We conducted a propensity score matching study and obtained 2 different paired groups of 58 patients with similar baseline risk factors, comorbidities, and risk scores who received DOACs or were treated with LAAC. The outcomes of the therapeutic strategy used (DOACs or LAAC) were assessed using the Cox regression analysis.

Results: During a median follow-up of 2.0 years [range 0.9-3.5] no differences regarding the primary endpoint (a composite of death, major bleeding, and stroke) were found (HR, 1.05; 95%CI, 0.15-7.51). Bleeding events were similar in both groups with no statistically significant differences being reported (HR, 1.79; 95%CI, 0.73-4.41). Mortality rate was numerically higher in patients on DOACs (31.8%) vs LAAC (26.4%). However, this finding did not reach statistical significance (HR, 0.70; 95%CI, 0.33-1.47; P = .343).

Conclusions: Compared to DOACs, LAAC has not shown any differences regarding embolic events, bleeding, and mortality in a population of elderly patients > 80 years of age. In our population, LAAC is a strategy as safe and effective as DOACs, and is an alternative to be taken into consideration in real-world patients > 80 years.

Keywords: Atrial fibrillation. Left atrial appendage closure. Direct oral anticoagulants. Embolic risk. Bleeding risk.


Introducción y objetivos: Existe poca información comparativa entre el cierre de la orejuela izquierda (COI) y los anticoagulantes orales de acción directa (ACOD). Nuestro objetivo fue comparar los resultados clínicos entre el COI y los ACOD en una población de pacientes mayores de 80 años.

Métodos: Se analizaron 1.144 pacientes octogenarios con fibrilación auricular provenientes de 3 hospitales terciarios. De ellos, 970 recibían ACOD y 174 fueron sometidos a COI. Ambos grupos presentaban similares factores de riesgo cardiovascular. El grupo de COI tenía mayor porcentaje de antecedentes de hemorragia, anemia y cáncer previo. Se llevó a cabo un análisis emparejado y se obtuvieron 2 grupos de 58 pacientes con similares factores de riesgo, comorbilidad y escalas de riesgo que fueron sometidos a COI o recibían tratamiento con ACOD. Los resultados de acuerdo con la estrategia terapéutica se obtuvieron mediante regresión de Cox.

Resultados: Durante una mediana de seguimiento de 2 años [rango: 0,9-3,5] no hubo diferencias en cuanto al evento combinado primario de muerte, hemorragia mayor o ictus (HR = 1,05; IC95%, 0,15-7,51). Las hemorragias fueron similares en ambos grupos, sin diferencias estadísticamente significativas (HR = 1,79; IC95%, 0,73-4,41). La mortalidad fue mayor en los pacientes con ACOD (31,8%) frente a aquellos con COI (26,4%), sin diferencias significativas (HR = 0,70; IC95%, 0,33-1,47).

Conclusiones: En comparación con los ACOD, el COI no ha mostrado diferencias en cuanto a eventos embólicos, hemorragias y mortalidad en una población de pacientes de edad avanzada. En nuestra cohorte, el COI es una alternativa que puede considerarse para los pacientes mayores de 80 años.

Palabras clave: Fibrilación auricular. Cierre de orejuela izquierda. Anticoagulantes orales directos. Riesgo embólico. Riesgo hemorrágico.

Abbreviations AF: atrial fibrillation. DOACs: direct oral anticoagulants. LAAC: left atrial appendage closure.


Atrial fibrillation (AF) has emerged as a clinically relevant issue of public health since it is associated with significant mortality and morbidity rates.1 AF is known to be a powerful risk factor for stroke independently increasing up to 5-fold across all ages. A total of 23.5% of all strokes occurred at 80-89 years of age are due to AF.2 The prevalence of AF is predicted to rise within the next few decades because of the growing population of elderly patients. Unfortunately, these patients are not often given oral anticoagulants. Only 35% of the patients aged ≥ 85 years without any clear contraindications for anticoagulation therapy receive the prescription.3 The reasons could be the increased risk of bleedings, especially intracranial and fatal bleedings,4 and also the frailty status of these patients. Since 2011, direct oral anticoagulants (DOACs) have shown a better risk-benefit ratio in patients with AF confirmed by a lower rate of stroke, intracranial hemorrhage, and mortality compared to warfarin.5 Still, with an improved safety and efficacy profile DOACs still present several shortcomings. The rate of discontinuation, the persistent risk of bleeding in high-risk populations or the risk of stroke when prescribed at a lower than recommended dosage are a matter of concern.6

Left atrial appendage closure (LAAC) was developed as an alternative to warfarin therapy in patients with AF. Several randomized controlled trials and few large registries have addressed the safety and efficacy profile of this technique.7,8 Recently, the evidence provided by long-term follow-up registries confirm that efficacy has similar endpoint rates compared to randomized controlled trials, and lower rates of stroke compared to the rates expected in untreated patients of similar risk.9

However, to this date, information comparing LAAC to DOACs therapy is scarce,10 and no comparison between both alternatives has been conducted in the elderly population. The aim of our study was to compare the clinical outcomes between LAAC and DOACs of an elderly population (> 80 years of age) using a propensity score matching study.


Study population

This retrospective multicenter study included a cohort of 1144 consecutive octogenarian patients with non-valvular AF treated with DOACs (N = 970) or LAAC (N = 174) from January 2014 through December 2018 at 3 Spanish and Canadian hospitals (Hospital Álvaro Cunqueiro, Vigo, Spain, Hospital Universitario, Salamanca, Spain, and Institut Universitaire de Cardiologie et Pneumologie de Quebec, Canada).

Authors defined non-valvular AF as AF unrelated to rheumatic mitral stenosis or prosthetic mechanical heart valves.11 Because the goal of the trial was to evaluate LAAC compared to DOACs in patients with non-valvular AF, patients treated with LAAC who received postoperative oral anticoagulation were not included in the study.

All the patients treated with LAAC were discussed and approved for LAAC by a multidisciplinary team. Regarding anticoagulated patients, the optimal dose of DOACs was based on the European recommendations.12 Electronic medical records were reviewed in all the patients to collect data regarding the baseline clinical variables, the therapeutic strategy, and the events occurred at the follow-up. The CHA2DS2-VASc and HAS-BLED scores were estimated for each patient.

The study was conducted in full compliance with the principles established in the Declaration of Helsinki and approved by the local ethics committee. Due to the retrospective nature of the study and its general interest, it was approved by each center local ethics committee without the need for informed consent.

Follow-up and outcomes

Primary endpoint was a composite of death, major bleeding, and stroke. Primary efficacy endpoints were all-cause mortality, and embolic events. Primary safety endpoint was the risk of major bleeding. Outcomes were censored at the last medical contact site in primary or secondary care, which was censored in November 2019 or until the end of anticoagulant therapy in the case of the DOAC group or the beginning of such therapy in the case of the LAAC group.

Embolic events were defined as a composite of any ischemic stroke, pulmonary embolism or peripheral embolism. Ischemic stroke was confirmed through concomitant imaging studies of the brain including computed tomography scan or magnetic resonance imaging. Major bleeding (MB) was defined using the definition established by the International Society on Thrombosis and Hemostasis.13 Bleeding was divided into intracranial hemorrhage (ICH) and non-ICH.

Statistical analyses

All statistical analyses were performed using IBM SPSS Statistics 25.0 and Stata 15.1 statistical software packages. Continuous variables were expressed as mean ± standard deviation and compared using the chi-square test. Categorical variables were expressed as percentages and compared using the Student t test.

A Cox analysis was performed to evaluate the unadjusted impact of LAAC vs DOAC on mortality, embolic and bleeding events. Due to the important differences reported in the baseline characteristics of patients treated with LAAC compared to those treated with DOACs we complemented our analysis with a propensity score matching (PSM) study. Patients were matched on a 1:1 ratio based on their nutritional status and on the propensity score using a < 0.2 caliper. Propensity score was estimated through logistic regression with the therapeutic group (LAAC or DOAC) as the dependent outcome with 21 baseline characteristics (table 1) as the independent variables. After PSM, we identified 58 patient-pairs with balanced baseline characteristics and no significant differences (table 2). Estimates were reported as hazard ratios (HR) with their 95% confidence intervals (95%CI). P values < .05 were considered statistically significant. Kaplan-Meier estimates were used to graphically evaluate the rate and timing of the events according to the therapeutic group (LAAC vs DOAC).

Table 1. Comparison of baseline characteristics between patients treated with DOACs or LAAC

Variables DOACs (N = 970) LAAC (N = 174) P
Age (years) 87.6 ± 3.6 83.6 ± 2.7 < .001
Female sex (%) 67.3 41.4 < .001
Body mass index (kg/m2) 29.1 ± 4.7 26.9 ± 3.7 < .001
Cardiovascular risk factors
 Hypertension (%) 70.3 90.2 < .001
 Diabetes (%) 20.3 37.4 < .001
Cardiovascular history
 Peripheral arterial disease (%) 12.2 29.9 < .001
 Ischemic heart disease (%) 14.4 35.1 < .001
 Previous heart failure (%) 25.8 42.0 < .001
 Previous embolic events (%) 26.4 17.5 .006
 Previous bleeding (%) 10.1 57.5 < .001
 Anemia (%) 27.2 69.5 < .001
 COPD (%) 8.7 20.1 < .001
 Dementia (%) 5.1 7.5 < .001
 Previous cancer (%) 8.7 22.4 < .001
Laboratory data
 Creatinine (mg/dL) 1.0 ± 0.3 1.4 ± 0.9 < .001
Echocardiographic data
 LVEF < 40% (%) 5.3 9.2 .042
 Severe aortic stenosis (%) 3.8 6.3 .129
Concomitant therapy
 Chronic use of NSAIDs (%) 4.6 14.9 < .001
 PPI (%) 50.3 86.2 < .001
Risk scores
 CHA2DS2-VASc (points) 4.3 ± 1.3 5.2 ± 1.3 < .001
 HAS-BLED (points) 2.5 ± 0.9 3.5 ± 0.8 < .001

COPD, chronic obstructive pulmonary disease; DOAC, direct oral anticoagulant; LAA, left atrial appendage; LVEF, left ventricular ejection fraction; NSAID, non-steroidal anti-inflammatory drug; PPI, proton pump inhibitor.


Baseline characteristics

Out of a total cohort of 1144 patients with AF, 970 patients were treated with DOACs while 174 underwent successful LAAC. The baseline clinical characteristics of the 2 groups (unmatched population) are shown on table 1. Patients from the DOACs group were slightly older being women more predominant. Previous history of bleeding was more common in patients treated with LAAC and the same thing happened with anemia, previous cancer, and dementia. Both groups had similar cardiovascular risk factors. Among the patients from the LAAC group less than 30% received dual antiplatelet therapy (27.6%), and 75.9% single antiplatelet therapy.

Regarding thrombotic and bleeding risk, the CHA2DS2-VASc and the HAS-BLED scores were significantly higher in the LAAC group (5.2 ± 1.3 vs 4.3 ± 1.3 for CHA2DS2VASc, and 3.5 ± 0.8 vs 2.5 ± 0.9 for HAS-BLED).

Clinical outcomes

Entire population

The median follow-up was 2.0 years [range 0.9-3.5]. The events shown on table 3 section “before PSM” we collected and analyzed at the follow-up. Embolic events tend to be more frequent among patients on DOACs without statistical significance. Major bleeding events were statistically significant in patients treate with LAAC compared to DOAC (P < .001).

Based on the univariate analysis, LAAC was associated with a higher rate of death, major bleeding, and stroke compared to DOACs (HR, 1.54; 95%CI, 1.06-2.24; P = .024). Regarding the efficacy endpoint of all-cause mortality and embolic events no significant differences were observed between both groups (HR, 0.87; 95%CI, 0.53-1.44). The same thing happened with embolic events and stroke (HR, 0.59; 95%CI, 0.26-1.36, and HR, 0.82; 95%CI, 0.33-2.08, respectively). Major bleeding was significantly higher in the LAAC group (HR, 3.43; 95%CI, 2.05-5.76) based on the univariate analysis. ICH did not differ between LAAC and DOACs (HR, 1.49; 95%CI, 0.43-5.19). Based on the univariate analysis, the all-cause mortality rate was not statistically significance (HR, 1.09; 95%CI, 0.80-1.50).

Propensity score matching study

After PSM a total of 58 patients were obtained in each group. The 2 groups were uniform regarding age (85.8 ± 3.7 vs 85.6 ± 2.5 years, P = .758), cardiovascular risk factors (32.8% vs 29.3% diabetes, P = .688; 81.0% vs 87.9% hypertension, P = .305), previous heart failure (37.9% vs 43.1% P = .570), creatinine levels (1.2 ± 0.5 mg/dL vs 1.2 ± 0.6mg/dL P = .809), and ischemic and bleeding risk (CHA2DS2-VASc, 4.7 ± 1.5 vs 4.7 ± 1.1; P = .834, and HAS-BLED, 3.2 ± 1.0 vs 3.2 ± 0.7 P = .834) as shown on table 2.

Table 2. Comparison of baseline characteristics after propensity score matching between patients treated with DOACs or LAAC

Variables DOACs (N = 58) LAAC (N = 58) P SMD
Age (years) 85.8 ± 3.7 85.6 ± 2.5 .758 -0.068
Female sex (%) 46.6 44.8 .852 0.035
Body mass index (kg/m2) 27.4 ± 4.2 27.8 ± 4.4 .667 0.092
Cardiovascular risk factors
 Hypertension (%) 81.0 87.9 .305 0.232
 Diabetes (%) 32.8 29.3 .688 -0.071
Cardiovascular history
 Peripheral arterial disease (%) 13.8 13.8 1.000 0.000
 Ischemic heart disease (%) 24.1 13.8 .155 -0.216
 Previous heart failure (%) 37.9 34.5 .699 -0.070
 Previous embolic events (%) 27.6 24.1 .672 -0.078
 Previous bleeding (%) 37.9 43.1 .570 0.104
 Anemia (%) 70.7 62.1 .326 -0.187
 COPD (%) 12.1 17.2 .431 0.129
 Dementia (%) 3.4 8.6 .242 0.288
 Previous cancer (%) 13.8 10.3 .569 -0.082
Laboratory data
 Creatinine (mg/dL) 1.2 ±0.5 1.2 ± 0.6 .809 -0.028
Echocardiographic data
 LVEF < 40% (%) 6.9 8.6 .729 0.059
 Severe aortic stenosis (%) 5.2 5.2 1.000 0.000
Concomitant therapy
 Chronic use of NSAIDs (%) 10.3 5.2 .298 -0.246
 PPI (%) 79.3 77.6 .821 -0.050
Risk scores
 CHA2DS2-VASc (points) 4.7 ± 1.5 4.7 ± 1.1 .834 -0.040
 HAS-BLED (points) 3.2 ± 1.0 3.2 ± 0.7 .826 -0.043

COPD, chronic obstructive pulmonary disease; DOAC, direct oral anticoagulant; LAA, left atrial appendage; LVEF, left ventricular ejection fraction; NSAID, non-steroidal anti-inflammatory drug; PPI, proton pump inhibitor; SMD, standardized mean difference.

At the follow-up, events were collected for both groups (see table 3 section “after PSM”). Patients on DOACs had more cardiovascular mortality compared to patients treated with LAAC (P = .555). Major bleeding was higher in patients treated with LAAC without statistical significance (P = .056).

Table 3. Clinical events in patients treated with DOAC and LAAC before and after propensity score matching between groups

Before PSM
Event DOAC (N = 970) LAAC (N = 174) P
No. Incidence rate (per 100 person/years) No. Incidence rate (per 100 person/years)
Mortality 308 13.5 (12.1-15.1) 46 14.7 (11.0-19.7) .543
CV mortality 96 4.2 (3.5-5.2) 9 2.9 (1.5-5.5) .617
Ischemic stroke 47 2.1 (1.6-2.8) 5 1.6 (0.7-3.9) .248
TIA 26 1.2 (0.8-1.7) 1 0.6 (0.1-4.3) .460
Peripheral embolism 2 0.1 (0.0-0.3) 0 - -
ICH 14 0.6 (0.4-1.0) 3 0.9 (0.3-2.9) .905
Major bleeding 48 2.1 (1.6-2.8) 21 7.5 (4.9-11.4) < .001
Minor bleeding 54 2.5 (1.9-3.2) 12 4.0 (2.3-7.0) .318
After PSM
Event DOAC (N = 58) LAAC (N = 58) P
No. Incidence rate (per 100 person/years) No. Incidence rate (per 100 person/years)
Mortality 20 15.2 (9.8-23.5) 12 11.2 (6.4-19.7) .343
CV mortality 6 4.5 (2.0-10.1) 3 2.8 (0.9-8.7) .555
Ischemic stroke 1 0.8 (0.1-5.5) 2 1.9 (0.5-7.7) .547
TIA 1 0.8 (0.1-5.5) 0 - -
Peripheral embolism 0 - 0 - -
ICH 2 1.6 (0.4-6.2) 1 0.9 (0.1-6.6) -
Major bleeding 4 3.0 (1.1-8.1) 9 9.2 (4.8-17.7) .056
Minor bleeding 6 4.6 (2.1-10.3) 2 1.9 (0.5-7.5) .292

CV, cardiovascular; DOAC, direct oral anticoagulant; ICH, intracranial hemorrhage; LAAC, left atrial appendage closure; PSM, propensity score matching; TIA, transient ischemic attack.

Regarding to primary endpoint (death, major bleeding, and stroke) after PSM, LAAC had a higher risk rate (HR, 1.62; 95%CI, 0.62-3.65) compared with DOACs. The primary efficacy endpoint (all-cause mortality, and embolic events) did not differ between both groups (HR, 0.83; 95%CI, 0.29-2.35).

No differences regarding embolic events were apparent between the 2 matched groups (HR, 1.05; 95%CI, 0.15-7.51) (figure 1). No statistically significant differences were found regarding the ischemic stroke (HR, 2.12; 95%CI, 0.19-23.39).

Figure 1. Analysis of embolic events at the follow-up between matched groups. 95%CI, 95% confidence interval; DOAC, direct oral anticoagulant; HR, hazard ratio; LAAC, left atrial appendage closure.

Safety endpoint (major bleeding) did not differ in either group (HR, 1.79; 95%CI, 0.73-4.41) (figure 2) after PSM.Also, ICH did not differ in either one of the 2 categories (HR, 0.61; 95%CI, 0.05-6.78).

Figure 2. Analysis of major bleeding events at the follow-up between matched groups. 95%CI, 95% confidence interval; DOAC, direct oral anticoagulant; HR, hazard ratio; LAAC, left atrial appendage closure.

Mortality rate was numerically higher in patients on DOACs. After the PSM study, this finding did not reach statistical significance (HR, 0.70; 95%CI, 0.33-1.47) (figure 3).

Figure 3. Analysis of mortality at the follow-up between matched groups. 95%CI, 95% confidence interval; DOAC, direct oral anticoagulant; HR, hazard ratio; LAAC, left atrial appendage closure.


This study has been designed with the intent to compare LAAC to DOACs in an elderly population (> 80 years old). The main finding of our study is that after PSM both DOACs and LAAC groups proved to have similar outcomes regarding the efficacy and safety profile.

As far as we are concerned this is the first study to compare both strategies in this population. We selected the cut-off value of 80 years not only because age is a known risk factor for stroke,2 but also because age is associated with bleeding events and fewer prescriptions of anticoagulants.14

Many studies have evaluated clinical outcomes with different antithrombotic strategies in elderly patients with AF. Two studies15,16 compared warfarin with aspirin supporting the use of anticoagulation in elderly and very elderly patients. Nonetheless, therapy with vitamin K antagonists is under-implemented in this population mostly due to the risk of falling (26.7%), poor prognosis (19.3%), bleeding history (17.1%), participant or family refusal (14.9%), older age (11.0%), and dementia (9.4%).17

As it has been discussed, DOACs provided an alternative to vitamin K antagonists. Dabigatran in both doses compared with warfarin—in patients aged ≥ 75 years—was associated with a similar or higher risk of major non-intracranial bleeding.18 Similarly, rivaroxaban described higher major gastrointestinal bleeding rates among the elderly population with no significant interaction between age and treatment efficacy.19 Apixaban proved beneficial compared to warfarin reducing the rates of stroke and major bleeding in our target population.20 Finally, edoxaban also proved beneficial in very elderly patients regarding major bleeding.21

Clinical trials21-24 comparing DOACs to warfarin led to the current guideline recommendation of DOACs as first-line therapy even in the elderly population.11,25 However, DOACs may present several limitations in this type of patients. We know from clinical registries that approximately 1 in 7 patients with AF receive reduced doses of DOACs even though they never met the criteria for reduced doses.26 Interestingly, this finding is more common among the elderly population. The rates of adverse events were higher in off-label dosed patients (HR for all-cause mortality, 2.18 [1.57-3.02]; HR for stroke, 1.50 [0.77-2.94]).27 In this sense, we did not evaluate the dose of DOACs in our patients, but it is known from previous registries that almost one third of the patients received inappropriate doses.28 Another important issue with DOACs is compliance. Recent data from studies conducted in the UK revealed poorer compliance with DOACs due to the lack of routine monitoring and, in some cases, the twice-daily dosing regime.29 Non-compliance in this group revealed adverse outcomes including mortality and stroke.30 Third, frailty is of major concern among the elderly population receiving anticoagulant drugs. A prospective study in hospitalized elderly patients showed that frailty is associated with a higher mortality rate at admission and a 2-fold increased risk of death at 1 year, particularly in anticoagulated patients.31 The risk of falling is an important parameter of frailty. In a recent study of older adults with a history of falls and AF, the risk of ICH at the follow-up was 1.9 times higher.32

LAAC may be a recommended therapeutic alternative in patients with AF ineligible for long-term oral anticoagulation who need stroke and embolism prevention according to the last EHRA/EACPI consensus statement.33 The PROTECT AF and PREVAIL 5-year outcome data were combined in a meta-analysis,34 and proved that LAAC with the Watchman device is equivalent to warfarin in stroke prevention and requires additional decreases of major bleeding and mortality. The safety and efficacy profile of the Amplatzer Cardiac Plug was examined in a multicenter study8 showing high procedural success rates and favorable outcomes preventing AF related thromboembolism.

A subanalysis of the EWOLUTION registry including patients aged ≥ 85 years showed that LAAC is a safe and effective procedure in these patients without any differences compared to younger patients regarding the annual stroke rates (2.0 vs 2.5 in ≥ 85 and < 85, respectively).35

Notwithstanding the above, the information available on this strategy compared to DOACs is scarce. To this date, only 2 studies have addressed this issue. The PRAGUE-17 was a prospective, multicenter, randomized non-inferiority trial conducted by Osmancik et al. that tried to compare LAAC with DOACs in high risk patients with AF (CHA2DS2-VASc ≥ 3, and HAS-BLED ≥ 2).36 Patients were younger compared to our cohort, mean age was 73.4 ± 6.7 in the LAAC group and 73.2 ± 7.2 in the DOACs group. They had similar CHA2DS2VASC scores (4.7 ± 1.5) in both groups, also similar to our cohort of patients. LAAC was non-inferior to DOAC therapy regarding the composite clinical and bleeding events through a median follow-up of 20.8 months. The rates of stroke and transient ischemic attack, cardiac death, clinically significant bleeding, and nonprocedural clinically significant bleeding did not differ between the study arms. These findings are consistent with the results obtained by Godino et al.10 Compared to our data, they selected a younger population (mean age 74.2 ± 7.7 in the LAAC group compared to 77.7 ± 6.9 in the DOACs group) with similar CHA2DS2VASc scores (4.3 ± 1.5 and 4.8 ± 1.5 in the LAAC and DOACs groups, respectively). They found similar outcomes between the 2 groups after PSM regarding thromboembolic events, ischemic stroke, transients ischemic attack, systemic embolism, and acute myocardial infarction, which is consistent with our own conclusions. Looking at the bleeding events, DOACs did not show an increased risk of major bleeding. In our population, results are consistent with previous findings even though we were dealing with older patients. Despite not being anticoagulated patients, the LAAC group did not have fewer bleeding events. Our hypothesis is that maybe many of them were treated with antiplatelet therapy.

Our observations are consistent with the previous studies mentioned, which supports the use of LAAC as an alternative to DOACs among elderly patients.

Study limitations

Our study has several limitations. First, its observational retrospective nature. Second, although rigorous matching was performed with 21 variables to neutralize the different clinical profile of patients, we cannot exclude the influence of other uncollected variables. Third, after PSM we achieved 2 well-balanced groups—though with a small sample size—that could lead to the underestimation of events at the follow-up. Also, we only selected patients with successful LAAC.

Despite all these limitations, we presented interesting data based on a multicenter study of consecutive octogenarian patients with non-valvular AF treated with DOAC vs LAAC.


This multicenter observational study proves the safety and efficacy profile after LAAC, with no differences regarding embolic and bleeding events, and mortality compared to DOACs in a propensity-matched population of real-world elderly patients > 80 years successful treated with LAAC without complications.


None whatsoever.


All authors contributed to patient recruitment, data curation, and process of manuscript review. J. Rodés-Cabau, A. Íñiguez-Romo, S. Raposeiras-Roubín, and R. Estévez-Loureiro were responsible for the study design. S. Raposeiras-Roubín, and B. Caneiro-Queija conducted the statistical analysis. B. Caneiro-Queija, S. Raposeiras-Roubín, and R. Estévez-Loureiro were responsible for preparing the manuscript.


R. Estévez-Loureiro is proctor for Watchman and has received honoraria from Boston Scientific. I. Cruz-González is proctor for Watchman and LifeTech and has received honoraria from Boston Scientific and Abbott Vascular. Rodés-Cabau has received a research grant from Boston Scientific. The remaining authors declared no other conflicts of interest.


  • Patients who are often treated with LAAC tend to be poor candidates for anticoagulation. As a matter of fact, older patients are excluded from randomized clinical trials and are more prone to receive reduced doses of DOACs. We know from previous trials about the noninferiority of LAAC compared to DOACs.


  • There was no current information on real-world older populations receiving DOACs compared to LAAC.
  • Although our data come from a registry they reflect our routine clinical practice; in a comparable profile population of older patients, LAAC might be as safe and effective as DOACs.


1. Zoni-Berisso M, Lercari F, Carazza T, Domenicucci S. Epidemiology of atrial fibrillation: European perspective. Clin Epidemiol. 2014;6:213-220.

2. Benjamin EJ, Virani SS, Callaway CW, et al. Heart Disease and Stroke Statistics-2018 Update: A Report From the American Heart Association. Circulation. 2018;137:e67-e492.

3. Go AS, Hylek EM, Borowsky LH, Phillips KA, Selby JV, Singer DE. Warfarin use among ambulatory patients with nonvalvular atrial fibrillation: the anticoagulation and risk factors in atrial fibrillation (ATRIA) study. Ann Intern Med. 1999;131:927-934.

4. Silverio A, Di Maio M, Prota C, et al. Safety and efficacy of non-vitamin K antagonist oral anticoagulants in elderly patients with atrial fibrillation: systematic review and meta-analysis of 22 studies and 440 281 patients. Eur Heart J Cardiovasc Pharmacother. 2021;7:f20-f29.

5. Ruff CT, Giugliano RP, Braunwald E, et al. Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: a meta-analysis of randomised trials. Lancet. 2014;383:955-962.

6. Kachroo S, Hamilton M, Liu X, et al. Oral anticoagulant discontinuation in patients with nonvalvular atrial fibrillation. Am J Manag Care. 2016;22:e1-8.

7. Holmes DR, Reddy VY, Turi ZG, et al. Percutaneous closure of the left atrial appendage versus warfarin therapy for prevention of stroke in patients with atrial fibrillation: a randomised non-inferiority trial. Lancet. 2009;374:534-542.

8. Tzikas A, Shakir S, Gafoor S, et al. Left atrial appendage occlusion for stroke prevention in atrial fibrillation: multicentre experience with the AMPLATZER Cardiac Plug. EuroIntervention. 2016;11:1170-1179.

9. Holmes DR, Jr., Reddy VY, Gordon NT, et al. Long-Term Safety and Efficacy in Continued Access Left Atrial Appendage Closure Registries. J Am Coll Cardiol. 2019;74:2878-2889.

10. Godino C, Melillo F, Bellini B, et al. Percutaneous left atrial appendage closure compared to non-vitamin K oral anticoagulants in patients with non-valvular atrial fibrillation and high bleeding risk.EuroIntervention. 2020;15:1548-1554.

11. Kirchhof P, Benussi S, Kotecha D, et al. 2016 ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS. Eur Heart J. 2016;37:2893-2962.

12. Steffel J, Verhamme P, Potpara TS, et al. The 2018 European Heart Rhythm Association Practical Guide on the use of non-vitamin K antagonist oral anticoagulants in patients with atrial fibrillation. Eur Heart J. 2018;39:1330-1393.

13. Schulman S, Kearon C. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost. 2005;3:692-694.

14. Tulner LR, Van Campen JP, Kuper IM, et al. Reasons for undertreatment with oral anticoagulants in frail geriatric outpatients with atrial fibrillation: a prospective, descriptive study. Drugs Aging. 2010;27:39-50.

15. Mant J, Hobbs FD, Fletcher K, et al. Warfarin versus aspirin for stroke prevention in an elderly community population with atrial fibrillation (the Birmingham Atrial Fibrillation Treatment of the Aged Study, BAFTA): a randomised controlled trial. Lancet. 2007;370:493-503.

16. Patti G, Lucerna M, Pecen L, et al. Thromboembolic Risk, Bleeding Outcomes and Effect of Different Antithrombotic Strategies in Very Elderly Patients With Atrial Fibrillation: A Sub-Analysis From the PREFER in AF (PREvention oF Thromboembolic Events-European Registry in Atrial Fibrillation). J Am Heart Assoc. 2017;6:e005657.

17. Cavallari I, Patti G. Efficacy and safety of oral anticoagulation in elderly patients with atrial fibrillation. Anatol J Cardiol. 2018;19:67-71.

18. Eikelboom JW, Wallentin L, Connolly SJ, et al. Risk of bleeding with 2 doses of dabigatran compared with warfarin in older and younger patients with atrial fibrillation: an analysis of the randomized evaluation of long-term anticoagulant therapy (RE-LY) trial. Circulation. 2011;123:2363-2372.

19. Halperin JL, Hankey GJ, Wojdyla DM, et al. Efficacy and safety of rivaroxaban compared with warfarin among elderly patients with nonvalvular atrial fibrillation in the Rivaroxaban Once Daily, Oral, Direct Factor Xa Inhibition Compared With Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET AF). Circulation. 2014;130:138-146.

20. Halvorsen S, Atar D, Yang H, et al. Efficacy and safety of apixaban compared with warfarin according to age for stroke prevention in atrial fibrillation: observations from the ARISTOTLE trial. Eur Heart J. 2014;35:1864-1872.

21. Kato ET, Giugliano RP, Ruff CT, et al. Efficacy and Safety of Edoxaban in Elderly Patients With Atrial Fibrillation in the ENGAGE AF-TIMI 48 Trial. J Am Heart Assoc. 2016;5:e003432.

22. Connolly SJ, Ezekowitz MD, Yusuf S, et al. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med. 2009;361:1139-1151.

23. Granger CB, Alexander JH, McMurray JJ, et al. Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med. 2011;365:981-992.

24. Patel MR, Mahaffey KW, Garg J, et al. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med. 2011;365:883-891.

25. January CT, Wann LS, Calkins H, et al. 2019 AHA/ACC/HRS Focused Update of the 2014 AHA/ACC/HRS Guideline for the Management of Patients With Atrial Fibrillation: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Rhythm Society in Collaboration With the Society of Thoracic Surgeons. Circulation. 2019;140:e125-e151.

26. Steinberg BA, Shrader P, Pieper K, et al. Frequency and Outcomes of Reduced Dose Non-Vitamin K Antagonist Anticoagulants: Results From ORBIT-AF II (The Outcomes Registry for Better Informed Treatment of Atrial Fibrillation II). J Am Heart Assoc. 2018;7:e007633.

27. Steinberg BA, Shrader P, Thomas L, et al. Off-Label Dosing of Non-Vitamin K Antagonist Oral Anticoagulants and Adverse Outcomes: The ORBIT-AF II Registry. J Am Coll Cardiol. 2016;68:2597-2604.

28. Ruiz Ortiz M, Esteve-Pastor MA, Roldan I, Muniz J, Marin F, Anguita M. Prognostic impact of inappropriate doses of direct oral anticoagulants in clinical practice. Rev Esp Cardiol. 2020;73:329-330.

29. Burn J, Pirmohamed M. Direct oral anticoagulants versus warfarin: is new always better than the old? Open Heart. 2018;5:e000712.

30. Borne RT, O’Donnell C, Turakhia MP, et al. Adherence and outcomes to direct oral anticoagulants among patients with atrial fibrillation: findings from the veterans health administration. BMC Cardiovasc Disord. 2017;17:236.

31. Gullon A, Formiga F, Diez-Manglano J, et al. Influence of frailty on anticoagulant prescription and clinical outcomes after 1-year follow-up in hospitalised older patients with atrial fibrillation. Intern Emerg Med. 2019;14:59-69.

32. Gage BF, Birman-Deych E, Kerzner R, Radford MJ, Nilasena DS, Rich MW. Incidence of intracranial hemorrhage in patients with atrial fibrillation who are prone to fall. Am J Med. 2005;118:612-617.

33. Glikson M, Wolff R, Hindricks G, et al. EHRA/EAPCI expert consensus statement on catheter-based left atrial appendage occlusion - an update. EuroIntervention. 2020;15:1133-1180.

34. Reddy VY, Doshi SK, Kar S, et al. 5-Year Outcomes After Left Atrial Appendage Closure: From the PREVAIL and PROTECT AF Trials. J Am Coll Cardiol. 2017;70:2964-2975.

35. Cruz-Gonzalez I, Ince H, Kische S, et al. Left atrial appendage occlusion in patients older than 85 years. Safety and efficacy in the EWOLUTION registry. Rev Esp Cardiol. 2020;73:21-27.

36. Osmancik P, Herman D, Neuzil P, et al; PRAGUE-17 Trial Investigators. Left Atrial Appendage Closure Versus Direct Oral Anticoagulants in High-Risk Patients With Atrial Fibrillation. J Am Coll Cardiol. 2020;75:3122-3135.

* Corresponding author.

E-mail address: raposeiras26@hotmail.com (S. Raposeiras-Roubín).

  @b_caneiro@S_Raposeiras  @RodrigoEstvez1


Introduction and objectives: A better positioning of left atrial appendage closure (LAAC) requires assessment of its clinical benefits to reduce thromboembolic and bleeding events in a real-word population.

Methods: Single-center retrospective study of our consecutive LAAC activity for 9 years. Both the device success and procedural success were registered as well as the reduction of the expected rates of thromboembolic and major bleeding events.

Results: A total of 260 LAAC procedures were performed in a population with nonvalvular atrial fibrillation with CHA2DS2-VASc and HAS-BLED scores of 4.3 ± 1.6 and 3.7 ± 1.2, respectively. Procedural success was 98.8%, and the rate of serious adverse events within the first 7 days was 2.3%. At a median follow-up of 2.5 ± 1.9 years and an estimated population of 637.9 patients-year, the thromboembolic event rate was 1.4 per 100 patients-year (75.5% risk reduction) and the rate of major bleeding was 3.0 per 100 patients-year (58.5% risk reduction), which was significantly lower than anticipated. The thromboembolic and major bleeding events per 100 patients-year showed a lower tendency for patients with very long follow-up (over 4 years) compared to the remaining of the population (0.7 vs 2.0 with P = .17, and 1.7 vs 4.0 with P = .09, respectively).

Conclusions: In our population, the LAAC showed high procedural success and a low rate of periprocedural adverse events. LAAC induced a significant reduction in the rate of predicted thromboembolic and hemorrhagic events, and this reduction was maintained even at very long follow-up.

Keywords: Percutaneous closure. Arterial embolism. Cerebral ischemia.


Introducción y objetivos: Conocer el beneficio clínico del cierre percutáneo de la orejuela izquierda (OI) en nuestro medio; en concreto, la reducción de eventos tromboembólicos y hemorrágicos, que permitiría un mejor posicionamiento de esta intervención.

Métodos: Estudio retrospectivo que recoge la actividad del cierre de OI en un centro durante 9 años. Se registraron la tasa de éxito del dispositivo y del procedimiento, así como las tasas de eventos tromboembólicos y de hemorragia mayor.

Resultados: Se evaluaron 260 procedimientos de cierre de OI en una población con fibrilación auricular no valvular y puntuación en las escalas CHA2DS2-VASc de 4,3 ± 1,6 y HAS-BLED de 3,7 ± 1,2. El éxito del procedimiento fue del 98,8%, y la tasa de eventos adversos graves en los primeros 7 días fue del 2,3%. Con un seguimiento medio de 2,5 ± 1,9 años y una población de 637,9 pacientes-año, la tasa de eventos tromboembólicos fue de 1,4 por 100 pacientes-año (75,5% de reducción del riesgo) y la de hemorragia mayor fue de 3,0 por 100 pacientes-año (58,5% de reducción del riesgo), ambas significativamente menores que las predichas. Las tasas de eventos por 100 pacientes-año en los pacientes con seguimiento muy largo (más de 4 años) mostraron tendencia a ser menores que en el resto de la población (0,7 frente a 2,0, con p = 0,17, para evento tromboembólico, y 1,7 frente a 4,0, con p = 0,09, para hemorragia mayor).

Conclusiones: En nuestra población, el cierre de la OI mostró un elevado éxito del procedimiento y una baja tasa de eventos inmediatos. El cierre de la OI indujo una significativa reducción en la tasa prevista de eventos tromboembólicos y hemorrágicos, y dicha reducción se mantuvo a muy largo plazo.

Palabras clave: Cierre percutáneo. Embolia arterial. Isquemia cerebral.


Percutaneous left atrial appendage closure (LAAC) has been extensively studied in clinical trials. Despite the excellent results of efficacy and safety regarding the LAAC from randomized clinical trials,1 these studies are limited by their design, which is still not applicable to our routine clinical practice. Maybe this is the reason why in our setting, the LAAC program is still far from reaching its full potential.2 Without detriment to the current level and grade of clinical recommendation for the LAAC,3 the medical community will only gain confidence in this procedure when further studies are presented assessing its performance in our routine clinical practice.

The LAAC is a solid structural procedure that in Spain is only second to transcatheter aortic valve implantation (TAVI).4 The experience gained with the LAAC has moved the focus of attention from the early aspects of success and safety towards other issues still not properly addressed such the performance of LAAC reducing long-term cardiovascular events or its lingering benefits over time.

To this day, there are very few papers gathering the long-term experience gained with the LAAC with a median follow-up of 2.5 years.1,5-9 It is only from this long-term perspective that we will understand the value of a procedure largely based on the prophylaxis of the thromboembolic complications occurred during the patient’s life.

The objective of this study was to present our own experience in the follow-up of the population treated with LAAC from the beginning of this program to assess its overall performance and, especially, the reduction of long and very long-term thromboembolic and bleeding events.


Our study is a retrospective analysis of the LAAC activity developed consecutively in a teaching hospital from March 2011 through February 2020. This procedure was indicated by different large volume hospital units including the internal medicine, neurology, and cardiology units. Our unit has included the LAAC as a strategic program within our structural heart procedures.

Left atrial appendage closure: the procedure and the device implanted

All procedures were performed in an identical working setting (facilities and personnel). However, 3 different modalities were used: on the one hand, general anesthesia and conscious sedation, both with transesophageal ultrasound guidance, and a third modality with fluoroscopy guidance only while the patient remained awake.

Although at the beginning of our experience only general anesthesia was used, 2.5 years later the possibility of conscious sedation administered by our personnel started to become a reality; the criterion to choose between general anesthesia or conscious sedation was logistical due to the discretional participation of the anesthesiology unit in structural heart procedures. In both modalities, the type of probe used for the transesophageal ultrasound was the exact same one.

The protocol of conscious sedation consisted of sedoanalgesia through the IV administration of 50 mg of pethidine followed by a bolus of 0.5 mg/kg of propofol with slow infusion in 3 min. with continuous monitorization of saturation and hemodynamics. After the introduction of the transesophageal probe several boluses of 10 mg of IV propofol were administered on demand based on the patient’s discomfort or rejection.

The procedure guided by fluoroscopy only was spared for cases with absolute or relative contraindication for transesophageal ultrasound use (in our unit we do not have intracavitary ultrasound) and for patients considered very frail for anesthetic induction; however, it became a reality 4 years after we started our experience. In these patients, a coronary computed tomography angiography was recommended to assess the left atrial appendage and discard the presence of an inner thrombus; in any case, an angiography was performed via transseptal access through a pigtail catheter from the left atrial appendage ostium without selective cannulation to discard the presence of thrombus. After catheterizing the left atrial appendage, a 180º rotational acquisition was performed through the injection of contrast at a flow rate of 8 mL/s with a total of 48 mL; by doing this a 3D image of the left atrial appendage was obtained (software i-Pilot, Siemens, Germany) that fused with the real fluoroscopy.

The 2 most popular devices in the market today were used: the WATCHMAN device in its WATCHMAN 2.5 and WATCHMAN Flex versions; Boston Scientific, United States) and the AMPLATZER ACP/Amulet device (Abbott, United States); the LAmbre device (Lifetech Scientific, China) was implanted anecdotically. The selection of one or the other did not follow any clinical or anatomical criteria and the alternate use of both devices was well-balanced. Only in fluoroscopy-guided procedures the AMPLATZER Amulet device was preferential since its delivery criteria are basically fluoroscopic.

Performance of left atrial appendage closure and follow-up

The definitions were based on the Munich consensus document regarding the LAAC.10 Successful LAACs were defined as successful devices (successful implantation of the first device selected) and successful procedures (uneventful final successful implantation within the first 24 hours). The device was released after confirming the suitability of ultrasound and fluoroscopic parameters. In cases performed under fluoroscopy guidance, position and stability were assessed over the fusion imaging as well as the lack of uncovered lobes in the angiography.

Regarding treatment after the implant, there was no pre-specified criterion and the patient’s bleeding risk was adjusted. All patients were assessed using a thoracic ultrasound within the first 24 hours prior to hospital discharge. Adherence to the transesophageal ultrasound control 1.5 months after the procedure was very irregular.

Follow-up was conducted back in February 2020 by reviewing the Andalusian (Diraya) electronic health record system. The appearance and dates of the following events were registered: death and causes, ischemic stroke/systemic embolism, major bleeding (incapacitating and major hemorrhages), and medical therapy at the follow-up. The futility of the LAAC was defined as mortality rate due to non-cardiac causes reported within the first year.

The performance of the LAAC at the follow-up was assessed using the risk reduction rate of thromboembolic (ischemic stroke/systemic embolism) or bleeding events (major bleeding) while taking into account the risk estimates from the CHA2DS2-VASc11 and HASBLED scores,12 respectively.

A 4-year follow-up limit has been established to start taking about «very long evolution» since this was the follow-up period of the Protect AF clinical trial1 that confirmed the superiority regarding mortality of LAAC over anticoagulation.

Statistical analysis

The estimates were obtained using IBM SPSS v26.0 and Epidat 4.2 statistical software. Initially, a descriptive analysis of data was conducted by generating means and standard deviations of numerical variables, and frequency and percentage distributions of qualitative variables.

The comparison between the demographic and clinical quantitative variables was conducted using the ANOVA test after verifying the hypotheses of normality using the Shapiro-Wilks test; when significant differences were seen, multiple comparisons were conducted using the Bonferroni correction.

The comparison among the different qualitative variables was conducted using contingency tables and the chi-square test.

The comparison between event incidence rates was conducted using the Rothman index score and 95% confidence intervals (95%CI) were estimated using Rosner’s method.

Finally, Kaplan-Meier curves were generated and then compared using the log rank test.



The population studied included 260 patients with nonvalvular atrial fibrillation aged between 42 and 92 years old. The clinical characteristics of the population are shown on table 1.

Table 1. Clinical characteristics of the population

Age (years) 74.8 ± 8.1
Males 160 (61.5%)
Risk factors
 Arterial hypertension 238 (91.5%)
 Diabetes (types 1 and 2) 118 (45.4%)
 Smoking 93 (35.8%)
 Dyslipidemia 130 (50%)
Kidney disease 64 (24.6%)
Ischemic heart disease 89 (34.2%)
Previous stroke
 Ischemic stroke 38 (14.6%)
 Hemorrhagic stroke 57 (21.9%)
CHA2DS2-VASc 4.3 ± 1.6
CHA2DS2-VASc ≥ 4 176 (67.7%)
HAS-BLED 3.7 ± 1.2
HAS-BLED ≥ 3 222 (85.4%)

Data are expressed as no. (%) or median ± standard deviation.

The most common indication for LAAC was the absolute contraindication for anticoagulant therapy due to hemorrhagic events in 229 cases (88.1%) or high-risk (2 patients with brain tumors and 1 patient with an aortic dissection; 1.1%). In 28 patients (10.8%) indication was due to the inability to take oral anticoagulants due to different bleeding risks: in 13 due to rejection to anticoagulant therapy, in 6 due to previous psychiatric history that did not recommend it, in 3 due to higher risk of falling, in 3 due to poor control of the international normalized ratio, and in other 3 due to cardioembolic stroke yet despite the proper anticoagulant therapy.

When the LAAC was indicated, therapy was mostly anticoagulation (68.8% of the patients): vitamin K antagonists (83 cases, 31.9%), direct anticoagulants (71 cases, 27.3%) or dual therapy (anticoagulation and single antiplatelet therapy, 25 cases, 9.6%). Regarding patients who were not on anticoagulant therapy prior to the LAAC, 48 of them (18.5%) were on antiplatelet therapy and 33 (12.7%) did not use any antiplatelet/anticoagulant drugs.

While follow-up was being conducted (February 2020 or prior to the patient’s death), our population was being treated with absence of antiplatelet/anticoagulant therapy (51 patients, 19,9%), single antiplatelet therapy with acetylsalicylic acid (135 patients, 52.5%), single antiplatelet therapy with clopidogrel (51 patients, 19,9%), dual antiplatelet therapy (14 patients, 5.4%) or anticoagulation (6 patients, 3%) (figure 1).

Figure 1. Evolution of antithrombotic therapy prior to the left atrial appendage closure (LAAC) until the final follow-up (%).

Procedural characteristics

Procedures were performed mostly under general anesthesia and monitored under transesophageal ultrasound guidance (59.6%). Conscious sedation, also monitored under transesophageal ultrasound guidance, was performed in 27.3% of all procedures. Only 13.1% of all procedures were performed under fluoroscopy guidance only.

The most commonly used device was the WATCHMAN (142 patients, 54.6%) followed by the AMPLATZER ACP/Amulet (116 patients, 44.6%), and occasionally the LAmbre (2 patients, 0.8%). Given the extension of the follow-up period, 2 models of the WATCHMAN (generation 2.5 in 125 patients and WATCHMAN Flex in 17 patients) and 2 models of the AMPLATZER device (ACP in 16 patients and Amulet in 100 patients) were used.

The device success rate was 98.5% (failed in 4 patients). Failed cases were due to the device not meeting the sealing criteria for the left atrial appendage so it had to be recaptured; after choosing a different device (different size, and in 1 case, also a different model), the procedure ended satisfactorily.

Procedural success was 98.8%; 1 oropharyngeal hemorrhage due to traumatic intubation and 2 tamponades were the reason for the lack of success. Tamponades (0.77%) were due, in the first case, to a perforation of the left atrial appendage in the recapture maneuver of the WATCHMAN device; the second case, after 24 hours, was due to the perforation of the left pulmonary artery possibly eroded by the LAmbre device. These 2 patients had a good clinical progression, the first one after pericardiocentesis and the second one after surgery with pericardial patch interposition between the pulmonary artery and the left atrial appendage. No deaths, strokes, or systemic embolisms were reported during the procedure or within the first 24 hours.

The number of serious adverse events reported within the first week was 6 (2.3%) as shown on table 2.

Table 2. Serious adverse events within the first 7 days after the implant

Day Event Description Death
Procedure Hemorrhage Traumatic intubation for general anesthesia No
Procedure Tamponade Pericardiocentesis No
1 day Tamponade Perforation of pulmonary artery Surgery No
4 days Bronchial aspiration Bronchial aspiration while eating Yes
4 days Hemorrhage Upper gastrointestinal bleeding Yes
6 days Hemorrhage Upper gastrointestinal bleeding No

The comparative analysis between the results of the first 50 LAACs and the remaining ones give us a glimpse of the existence of a learning curve that can be seen in the procedural variables that assess the operator’s technical skills (significant reduction of fluoroscopy time and radiation dose from the first 50 procedures): 13.6 ± 5.5 min vs 18.7 ± 18.2 min and 18 413 µGym ± 11 622 µGym vs 24 798 µGym ± 18 802 µGym,2 respectively with P values = .03). However, no differences were found in the procedural success rate (98% for the first 50 cases and 99% for the remaining ones).

The procedural characteristics of the left atrial appendage closure are shown on table 3.

Table 3. Procedural characteristics

Procedural modality
 General anesthesia 155 (59.6%)
 Conscious sedation 71 (27.3%)
 Fluoroscopy 34 (13.1%)
 ACP-Amulet 116 (44.6%)
 WATCHMAN 142 (54.6%)
 LAmbre 2 (0.8%)
Device size (mm) 25.2 ± 3.4
Fluoroscopy time (min) 14.6 ± 9.7
Radiation (µGym2) 19 636 ± 13 488
Device success 98.5%
Procedural success 98.8%

Data are expressed as no. (%) or median ± standard deviation.

Follow-up and events

With a median follow-up of 2.5 years ± 1.9 years (median, 1.4 years; 95%CI, 1.1 to 1.9 years) our series included 637.9 patients-year.

A total of 58 deaths were reported at the follow-up (22.3% of the sample, 9.1% patients-year). Half of them were due to cardiac causes (4.6% patients-year). A total of 6 deaths were due to noncardiac causes within the first year, which means that LAAC futility rate was 2.3%.

Events such as ischemic strokes/systemic embolisms were reported in 9 patients (1.4% patients-year, 95%CI. 0.6-2.7); compared to the estimated risk of 5.7% patients-year, the reduction of relative risk was 75.2% (P < .001). A total of 19 major hemorrhages were reported (3.0% patients-year. 95%CI. 1.8-4.7), which is a 58.5% reduction of relative risk compared to the estimated risk of 7.2% patients-year (P < .001)

The assessment of the protective capacity of LAAC to avoid long-term thromboembolic phenomena and major hemorrhages is very relevant. Events were compared in patients with follow-ups of up to 4 years (N = 206; 346.7 patients-year) and in patients beyond this 4-year follow-up mark (N = 54; 291.3 patients-year). It was confirmed that, over time, protection against thromboembolic and hemorrhagic events still remains, and there is even a decreasing tendency: annual rate per 100 patients-year for ischemic stroke/embolism of 2.0 vs 0.7 (P = 0.17) and for major hemorrhages of 4.0 vs 1.7 (P = .09) in patients with up to 4-year follow-ups and longer follow-ups, respectively. The comparison of event-free survival rates for thromboembolism and major bleeding between the different populations based on the duration of the follow-up did not show any significant results (log rank with P = .10 for thromboembolisms and P = .54 for hemorrhages) (figure 2).

Figure 2. Thromboembolic event-free (A) and major bleeding-free (B) survival curves of patients with < 4 year (green) and > 4-year follow-up (gray). The curve comparison does not show any significant differences for either one of the events.

Follow-up based on the type of device implanted

A comparative analysis of the event-free survival rate in patients treated with the WATCHMAN and AMPLATZER devices found no significant differences between the 2 regarding their protective capabilities against ischemic strokes/systemic embolisms (log rank P = .86); however, the WATCHMAN showed a major hemorrhage-free cumulative incidence rate superior to the AMPLATZER device (log rank P = .01) (figure 3).

Figure 3. Thromboembolic event-free (A) and major bleeding-free (B) survival curves of patients with the AMPLATZER (green) and the WATCHMAN device (gray). No differences were seen in the devices used for thromboembolic events, but there were differences regarding major bleeding with a higher event-free survival rate in patients treated with the WATCHMAN device.


This real-world single-center registry shows our experience performing left atrial appendage closure in 260 consecutive patients with nonvalvular atrial fibrillation over the last 9 years. Results have been exposed in an attempt to answer the following questions: what were the results of LAAC in our population? what is the actual performance of LAAC reducing thromboembolic or hemorrhagic events compared to the estimated risk rates? and finally, is this this event reduction maintained at the follow-up?

The clinical characteristics of our population are consistent with those of the LAAC target population in the routine clinical practice. Thus, our population showed clinical characteristics of thromboembolic risk that were similar to those published in large registries: the CHA2DS2-VASc score of 4.3 was intermediate between the AMPLATZER Amulet registry13 with a CHA2DS2-VASc score of 4.2 and the NCDR registry14 with a CHA2DS2-VASc score of 4.6. Regarding the risk of bleeding, in our population the mean HAS-BLED score was 3.8, slightly higher compared to the numbers already published, and situated between the EWOLUTION registry with a HAS-BLED score of 2.315,16 and the AMPLATZER Amulet registry with a HAS-BLED score of 3.3.13

This high risk of bleeding of our population may be explained by the fact that the indication for LAAC for almost 90% of the patients was a past medical history of bleeding (mostly gastrointestinal followed by cerebral); for the remaining 10%, the indication for LAAC was the «inability to take oral anticoagulants due to different risks of bleeding»,10 that is, by a number of reasons that forced the patient (5% of the population) or the doctor to make the decision of choosing mechanical local therapy over the anticoagulant therapy. Although in our case the volume of elective decisions regarding the LAAC is far from the volume reported in the German registry LAARGE,17 where patient selection was essential to propose the indication in a fourth of the population, a reflection can be made on to what extent information brought to the patient is decisive to generalize this therapy.

The LAAC is a procedure with high device and procedural success rates in most of the series already published. In our series the device success rate in the implant was 98.5%; 1.5% of failed procedures were due to an erroneous selection of the size of the device. However, the left atrial appendages of all the patients from the series were eventually sealed, which contrasts with up to 7% of the procedures cancelled due to inaccessible left atrial appendage anatomies;14 this may have to do with our capacity to approach this procedure using different modalities (general anesthesia, conscious sedation, and fluoroscopy without ultrasound guidance) and different types of occluder devices (yet the device had to be changed for a different one only in 1 patient in order to finish the procedure). Our procedural success rate was 98.8%, which is higher compared to the rate reported in other registries with similar populations regarding their baseline clinical characteristics.14 Regarding procedural safety, our adverse event rate within the first 7 days after the procedure was 2.3%, which is consistent with the rate reported by large registries.13-16 Overall, this speaks of the progressive decline seen in the rate of adverse events reported during the early stage after the procedure.

In the consecutive analysis of procedural results like the left atrial appendage closure, right from the beginning of our experience and until today, the presence of a learning curve may be anticipated. However, beyond procedural variables like the radiation duration and dose, no differences were reported regarding the procedural success rate between the early period and the rest of the experience. Standardizing procedures and training the operators may be the reasons of the high success rate reported in left atrial appendage closure despite the poor early experience reported.18

Our registry, with a median follow-up of 2.5 years and a fifth of the patients with follow-up periods > 4 years allows us to assess the efficacy of the LAAC with a certain perspective. In the first place, mortality rate is surprisingly high since 22.3% of the patients included died at the follow-up. This is an annual mortality rate of 9.1%, 3 times higher compared to the 4-year follow-up of the Protect AF,1 but it is nearly identical as other registries with a similar risk population compared to ours.9 The highest mortality risk seen at the follow-up has been associated with factors such as age, male sex, history of stroke or intracranial hemorrhage, low ejection fraction, and chronic kidney disease;8,9 in any case, this high mortality rate seen at the follow-up shows how frail this diseased population really is, which would justify an interesting debate on the futility of the LAAC in some patients19 (2.3% in our series).

The primary endpoint of LAAC is to reduce the risk of cardiac embolism in a population with nonanticoagulated atrial fibrillation. In our case, the annual rate of ischemic stroke and embolism was 1.4%, which was a significant reduction of the relative risk of 75%, which is consistent with the best data reported in the medical literature.20 Regarding major bleeding, in a population with an estimated rate of bleeding > 7%, our rate was 3.0%, that is, half the rate reported by other authors.9

To this day, very few studies have been conducted on the long-term efficacy profile of the left atrial appendag-e closure. In the Ibérico II registry,8 the rate of thromboembolic events remained low while the rate of major bleeding was lower compared to the early rates at the 2-year follow-up. In our population, the analysis of patients with very long clinical courses (implantation times > 4 years) revealed that the efficacy of the LAAC still remains. Also, that thromboembolic and bleeding events showed a tendency towards a lower incidence rate compared to the earliest stage.


This study has some limitations. In the first place, no systematic antithrombotic pattern was followed after implantation. Instead, it was left to the operator’s discretion, which may have impacted the short-term bleeding rate. On the other hand, no systematic imaging follow-up was arranged 45 days to 3 months after implantation, which means that an important piece of information was lost: the rate of thrombosis associated with the device, lack of residual sealing…


In our setting, left atrial appendage closure is an effective therapy for patients with nonvalvular atrial fibrillation and coagulation issues. It significantly reduces the rates of thromboembolic and hemorrhagic events that remain consistent in the very long term.


No funding related for this work.


R.J. Ruiz-Salmerón: author of the manuscript; M. Ronquillo-Japón, C. Robles-Pérez, M. Iglesias-Blanco, C. Rubio-Iglesias,R. García de la Borbolla, C. Carrascosa-Rosillo, S. Rodríguez de Leiras, M. Vizcaíno-Arellano, and I. Méndez-Santos: critical review and J. Polo-Padillo: statistical analysis.


None reported.


  • – It is estimated that only 5% of the patients with nonvalvular atrial fibrillation and inability to use oral anticoagulant therapy have benefited from the left atrial appendage closure. The evidence from randomized clinical trials is based on a population that is not similar to the one considered eligible for LAAC in the real world, which is a limitation. Relevant real-world registries do not have long-term follow-ups either.


  • – Our study provides data on the performance of the left atrial appendage closure in our routine clinical practice on procedural success and performance reducing thromboembolic and major bleeding events, which, overall, is significant compared to the estimated rates and also remains consistent over the very long-term.


1. Reddy VY, Sievert H, Halperin J, et al. Percutaneous Left Atrial Appendage Closure vs Warfarin for Atrial Fibrillation:A Randomized Clinical Trial. JAMA. 2014;312:1988-1998.

2. Fukutomi M, De Backer O, Søndergaard L. Indications, current adoption and future perspectives for percutaneous left atrial appendage closure. EuroIntervention. 2019;14:1707-1709.

3. Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association of Cardio-Thoracic Surgery (EACTS). Eur Heart J. 2021;42:373-498.

4. Ojeda S, Romaguera R, Cruz-González I, Moreno R. Registro Español de Hemodinámica y Cardiología Intervencionista. XXIX Informe Oficial de la Sección de Hemodinámica y Cardiología Intervencionista de la Sociedad Española de Cardiología (1990-2019). Rev Esp Cardiol. 2020;73:927-936.

5. Wiebe J, Franke J, Lehn K, et al. Percutaneous Left Atrial Appendage Closure With the Watchman Device:Long-Term Results Up to 5 Years. J Am Coll Cardiol Intv. 2015;8:1915-1921.

6. Betts TR, Leo M, Panikker S, et al. Percutaneous left atrial appendage occlusion using different technologies in the United Kingdom:A multicenter registry. Catheter Cardiovasc Interv. 2017;89:484-492.

7. Korsholm K, Nielsen KM, Jensen JM, Jensen HK, Andersen G, Nielsen-Kudsk JE. Transcatheter left atrial appendage occlusion in patients with atrial fibrillation and a high bleeding risk using aspirin alone for post-implant antithrombotic therapy. EuroIntervention. 2017;12:2075-2082.

8. López-Mínguez JR, Nogales-Asensio JM, Infante De Oliveira E, et al. Reducción de eventos a largo plazo tras el cierre de la orejuela izquierda. Resultados del Registro Ibérico II. Rev Esp Cardiol. 2019;72:449-551.

9. Regueiro A, Cruz-Gonzalez I, Bethencourt A, et al. Long-term outcomes following percutaneous left atrial appendage closure in patients with atrial fibrillation and contraindications to anticoagulation. J Interv Card Electrophysiol. 2018;52:53-59.

10. Tzikas A, Holmes DR, Gafoor S, et al. Percutaneous left atrial appendage occlusion:the Munich consensus document on definitions, endpoints and data collection requirements for clinical studies. EuroIntervention. 2016;12:103-111.

11. Lip GY, Nieuwlaat R, Pisters R, Lane DA, Crijns HJ. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based approach:the euro heart survey on atrial fibrillation. Chest. 2010;137:263-272.

12. Pisters R, Lane DA, Nieuwlaat R, de Vos CB, Crijns HJ, Lip GY. A novel user-friendly score (HAS-BLED) to assess one-year risk of major bleeding in atrial fibrillation patients:The Euro Heart Survey. Chest. 2010;138:1093-1100.

13. Landmesser U, Schmidt B, Nielsen-Kudsk JE, et al. Left atrial appendage occlusion with the AMPLATZER Amulet device:periprocedural and early clinical/echocardiographic data from a global prospective observational study. EuroIntervention. 2017;13:867-876.

14. Freeman JV, Varosy P, Price MJ, et al. The NCDR left atrial appendage occlusion registry. J Am Coll Cardiol. 2020;75:1503-1518.

15. Boersma LV, Schmidt B, Betts TR, et al. EWOLUTION investigators. Implant success and safety of left atrial appendage closure with the WATCHMAN device:peri-procedural outcomes from the EWOLUTION registry. Eur Heart J. 2016;37:2465-2474.

16. Boersma LV, Ince H, Kische S, et al. Efficacy and safety of left atrial appendage closure with WATCHMAN in patients with or without contraindication to oral anticoagulation:1-year follow-up outcome data of the EWOLUTION trial. Heart Rhythm. 2017;14:1302-1308.

17. Fastner C, Nienaber CA, Park JW, et al. Impact of left atrial appendage morphology on indication and procedural outcome after interventional occlusion:results from the prospective multicenter German LAARGE registry. EuroIntervention. 2018;14:151-157.

18. Sawant AC, Seibolt K, Sridhara S, et al. Operator experience and outcomes after transcatheter left atrial appendage occlusion with the Watchman device. Cardiovasc Revasc Med. 2020;21:467-472.

19. Asmarats L, Rodés-Cabau J. Resultados a largo plazo tras el cierre de la orejuela izquierda:ampliando la perspectiva en la prevención no farmacológica del ictus en pacientes con fibrilación auricular. Rev Esp Cardiol. 2019;72:440-442.

20. Glikson M, Wolff R, Hindricks G, et al. EHRA/EAPCI expert consensus statement on catheter-based left atrial appendage occlusion –an update. EuroIntervention. 2020;15:1133-1180.

* Corresponding author: Servicio Endovascular, Hospital Virgen Macarena, Avda. Doctor Fedriani 3, 41009 Seville, Spain.

E-mail addres: rjruizsalmeron@yahoo.es (R.J. Ruiz-Salmerón).

Original articles

Recic 22 074 Uk F3

Recic 22 037 F3


Original articles

Recic 22 043 Uk F3


Scientific letters

Recic 22 059 Uk F1


Debate: Drug-coated balloons for de novo coronary artery lesions

Recic 22 106 Author
Recic 22 102 Author

Percutaneous treatment of partial anomalous pulmonary venous connection with dual drainage

Rafael González-Manzanares, Soledad Ojeda, et al.

How would I approach it?
Felipe Hernández Hernández

Case resolution
Rafael González-Manzanares, Soledad Ojeda, et al.

Cookies user preferences
We use cookies to ensure you to get the best experience on our website. If you decline the use of cookies, this website may not function as expected.
Accept all
Decline all
Read more
Tools used to analyze the data to measure the effectiveness of a website and to understand how it works.
Google Analytics